Are labor-controlled firms more innovative? Evidence from a regression discontinuity design

We examine how labor representation on corporate boards affects innovation. Exploiting German co-determination law, which requires one-third (one-half) labor-elected directors in firms above 500 (2000) employees, we implement a regression discontinuity design using administrative employer-inventor-patent data. Crossing either threshold raises innovative productivity, measured as the forward citation-weighted number of patents per firm, by about 20%–25%. This increase is driven by more patents, while forward citations do not rise and decline significantly at the parity cutoff, indicating a shift toward more incremental, less path-breaking inventions. Additional evidence shows that labor-controlled firms insure workers against adverse labor market shocks, lengthen inventors’ employment spells, and offer wage incentives that reward patent quantity rather than quality. Overall, labor representation expands the number of patented inventions but dilutes their average citation impact, reallocating innovative effort in labor-controlled firms from path-breaking to incremental projects. • German co-determination law mandates one-third (one-half) labor-elected directors in firms above 500 (2000) employees, creating two regression discontinuities for causal identification. • Crossing either threshold raises innovative productivity, measured as forward citation- weighted patents per firm, by approximately 20%–25%. • The increase is driven by more patents (exploitative innovation), while patent gen- erality (exploratory innovation) does not rise and declines significantly at the parity cutoff. • Labor-controlled firms insure inventors against adverse labor market shocks, lengthen employment spells, and reduce the intensive margin of labor supply. • Inventors in labor-controlled firms earn similar marginal income for high- and low- quality patents, providing no financial incentive to pursue path-breaking over incre- mental inventions.